coreyhaines31/makerskills / skills/radar
radar
FlaggedWhen you want to monitor known sources on a schedule and feed the good stuff into your second brain. Configure sources once (YouTube, RSS/newsletters, subreddits, Hacker News, Bluesky, Mastodon, X, LinkedIn, keyword searches); each run fetches only what's new, scores it against your stated focus, writes one digest to the vault, and auto-captures high-signal items into raw/. Everything else waits in the digest until you promote it. Modes — run, digest, promote, add / sources / pause, doctor, schedule (daily launchd or systemd job). Triggers on "/radar," "run my radar," "check my sources," "what's new from my sources," "add a source," "monitor this channel," "watch this subreddit," "track this account," "daily digest," "promote item 4," "radar doctor." Complements second-brain (radar fills raw/, second-brain compiles it) and deep-research (standing surveillance vs a one-off dive).
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 15.9 KB |
| references/fetchers.md | markdown | 22.7 KB |
| references/scheduling.md | markdown | 7.5 KB |
| references/sources-schema.md | markdown | 6.2 KB |
| references/templates/digest.md | markdown | 1.1 KB |
| references/templates/sources.example.yaml | config | 4.5 KB |
references/scheduling.md
View on GitHub# Scheduling — daily unattended runs
Goal: a headless agent run of `/radar run` fires every morning, writes the digest, commits the vault, and leaves a log you can read when it doesn't.
The examples use Claude Code's headless mode (`claude -p`). With another agent, swap in its non-interactive command (e.g. `codex exec "run the radar skill in run mode"`) everywhere `claude -p` appears. The permission notes below are Claude Code's; check your agent's equivalent.
- **macOS**: launchd (below).
- **Linux**: a systemd user timer (see "Linux" near the end), or cron.
- **Anything else, or conditional schedules**: `loopify`.
## macOS (launchd)
Use launchd, not cron. On macOS, cron still works but launchd handles sleep/wake properly — `StartCalendarInterval` runs the job when the machine wakes if it was asleep at the scheduled time, which is exactly the laptop case.
## The plist
`~/Library/LaunchAgents/com.makerskills.radar.plist` (replace `/Users/<you>` with your home directory; launchd doesn't expand `~` or `$HOME` outside the shell command):
```xml
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
<key>Label</key>
<string>com.makerskills.radar</string>
<key>ProgramArguments</key>
<array>
<string>/bin/zsh</string>
<string>-lc</string>
<string>$HOME/.config/makerskills/radar/run.sh</string>
</array>
<key>WorkingDirectory</key>
<string>/Users/<you></string>
<key>EnvironmentVariables</key>
<dict>
<key>PATH</key>
<string>/Users/<you>/.local/bin:/opt/homebrew/bin:/usr/bin:/bin:/usr/sbin:/sbin</string>
<key>SECOND_BRAIN_VAULT</key>
<string>/Users/<you>/<path-to-your-vault></string>
<key>MAKERSKILLS_CONFIG</key>
<string>/Users/<you>/.config/makerskills</string>
</dict>
<key>StartCalendarInterval</key>
<dict>
<key>Hour</key><integer>7</integer>
<key>Minute</key><integer>0</integer>
</dict>
<key>StandardOutPath</key>
<string>/Users/<you>/.config/makerskills/radar/logs/stdout.log</string>
<key>StandardErrorPath</key>
<string>/Users/<you>/.config/makerskills/radar/logs/stderr.log</string>
<key>RunAtLoad</key>
<false/>
</dict>
</plist>
```
## Install
```bash
mkdir -p ~/.config/makerskills/radar/logs
launchctl unload ~/Library/LaunchAgents/com.makerskills.radar.plist 2>/dev/null
launchctl load ~/Library/LaunchAgents/com.makerskills.radar.plist
launchctl list | grep com.makerskills.radar # → PID/exit-status label
```
Run it once by hand before trusting the schedule:
```bash
launchctl start com.makerskills.radar
tail -f ~/.config/makerskills/radar/logs/stdout.log
```
## The six ways this fails silently
Every one of these produces "the job ran and nothing happened," so check them in this order:
1. **`PATH`.** launchd jobs get a minimal `PATH` — not your shell's. `claude` lives in `~/.local/bin` and Homebrew tools in `/opt/homebrew/bin`; neither is on the default. Hence `-lc` (loads the login profile) **and** an explicit `PATH` in `EnvironmentVariables`. Belt and braces, deliberately.
2. **Run under the shell that has your keys — this one is silent and expensive.** If API keys and vault paths live in `~/.zshenv`, **only zsh reads it**. Under `bash -lc` they simply aren't there, so the job runs keyless: X and LinkedIn sources degrade every morning while the identical command works perfectly when you test it in your terminal. Verified:
```
$ zsh -c 'echo ${SCRAPECREATORS_API_KEY:+set}' → set
$ bash -lc 'echo ${SCRAPECREATORS_API_KEY:+set}' → (empty)
```
`zsh -lc` inherits everything with no secrets duplicated into the plist — which matters, since a plist is world-readable plaintext and an API key has no business in one. The plist still sets `SECOND_BRAIN_VAULT` and `MAKERSKILLS_CONFIG` as a floor for the non-secret paths.
3. **Permissions.** An unattended `claude -p` can't answer a permission prompt — it will hang or bail. `--permission-mode acceptEdits` covers file writes; make sure the Bash commands radar needs (`curl`, `git`, `yt-dlp`) are already allowlisted in settings, or the run stalls on the first fetch.
4. **Credentials.** Even under zsh, confirm the job actually sees the key — `doctor` reports which keys resolved and from where. If you later move keys to the Keychain, a launchd agent may need a one-time "Always Allow" on the item's ACL.
5. **Git auth.** The vault push needs credentials available non-interactively. With `gh auth` or an SSH key in the login keychain this works; with a passphrase-protected key and no agent loaded, the push hangs. Test with `launchctl start` before trusting it — an interactive test in your terminal proves nothing about the launchd environment.
6. **Full Disk Access.** A vault under `~/Documents`, `~/Desktop`, or iCloud Drive is a protected location; if the job can read `~/.config` and not the vault, this is why. Grant FDA to `/bin/zsh` or move to a wrapper script with FDA.
## Wrapper script
Putting the command in a script instead of inline in the plist makes it editable without an unload/load cycle, and gives you somewhere to put a run marker:
```bash
#!/bin/zsh
# ~/.config/makerskills/radar/run.sh
# zsh, not bash: ~/.zshenv carries SCRAPECREATORS_API_KEY and friends.
set -uo pipefail
export PATH="$HOME/.local/bin:/opt/homebrew/bin:$PATH"
export SECOND_BRAIN_VAULT="$HOME/<path-to-your-vault>"
export MAKERSKILLS_CONFIG="$HOME/.config/makerskills"
LOG="$MAKERSKILLS_CONFIG/radar/logs/$(date +%F).log"
echo "=== radar run $(date -Iseconds) ===" >> "$LOG"
claude -p "/radar run" --permission-mode acceptEdits >> "$LOG" 2>&1
echo "=== exit $? ===" >> "$LOG"
```
The plist above already points at this script. Keep the shebang matched to the shell that holds your keys (`#!/bin/zsh` for `~/.zshenv`).
Add a credential assertion near the top so a keyless run fails loudly instead of quietly degrading every source:
```bash
[ -n "${SCRAPECREATORS_API_KEY:-}" ] || echo "WARN: no ScrapeCreators key — X/LinkedIn will degrade" >> "$LOG"
```
## Changing the schedule
- **Twice daily**: `StartCalendarInterval` accepts an *array* of dicts — one per firing time.
- **Weekdays only**: add `<key>Weekday</key><integer>1</integer>` … through 5, as separate dicts in the array.
- **Anything conditional or self-pacing** (run until X, back off when quiet, poll a metric): that's `loopify` territory, not launchd.
## Linux (systemd user timer)
`~/.config/systemd/user/radar.service`:
```ini
[Unit]
Description=radar daily run
[Service]
Type=oneshot
ExecStart=%h/.config/makerskills/radar/run.sh
```
`~/.config/systemd/user/radar.timer`:
```ini
[Unit]
Description=radar daily at 07:00
[Timer]
OnCalendar=*-*-* 07:00:00
Persistent=true
[Install]
WantedBy=timers.target
```
```bash
systemctl --user daemon-reload
systemctl --user enable --now radar.timer
systemctl --user start radar.service # run once by hand before trusting it
journalctl --user -u radar.service -n 50
```
`Persistent=true` is the Linux equivalent of launchd's run-on-wake. The same silent-failure list applies: systemd services get a minimal environment, so the wrapper script must set `PATH` and source whichever file holds your keys (use `#!/bin/bash` and source it explicitly if you don't use zsh). For runs while logged out, `loginctl enable-linger $USER`.
## Uninstall
```bash
# macOS
launchctl unload ~/Library/LaunchAgents/com.makerskills.radar.plist
rm ~/Library/LaunchAgents/com.makerskills.radar.plist
# Linux
systemctl --user disable --now radar.timer
```
State, sources, and past digests are untouched by this — the schedule is the only thing removed.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |